Delivery and integration
This page covers what every Pixalate data feed has in common: how you get access, how files reach you, how they’re named, and how to pick up new files on time. Each feed’s own page covers its schema and how to apply it.
Video (1.5 min): check, download and load a data feed file, and poll on schedule.
Get access
Data feeds are part of your Pixalate contract. To subscribe to a feed, add one or change how it’s delivered, contact your Pixalate account manager.
For FTP delivery, Pixalate gives you a username and password for ftp.pixalate.com. Each feed is in its own folder.
Some feeds have their own access terms. For example, High Risk Mobile Apps 2.0 Standard requires acceptance of updated terms, and the Enterprise version requires a subscription upgrade. See the feed’s page for details.
Delivery methods
FTP is available for every feed. Some feeds can also be delivered in other ways:
To use a method other than FTP, or to ask whether API access is available for a feed, contact your account manager.
Download over FTP
Each file’s last-modified time tells you when it was published. Check it before you download:
Then download the file:
Replace USERNAME, PASSWORD, FEED_FOLDER and FILE_NAME with the values from your Pixalate onboarding details and the feed’s naming convention.
Download over HTTPS (WebDAV)
The same server is also available over HTTPS with WebDAV. To list a folder, with each file’s metadata, send a PROPFIND request. The response is an XML document:
When you know the file’s path, download it:
To browse the files by hand on a Mac, open Finder, choose Go > Connect to Server, enter https://ftp.pixalate.com, and sign in with your username and password. Your folder opens on the desktop, and you can drag files out of it.
File format
Every feed is a CSV file. The columns differ by feed and are listed on each feed’s page. Some files have no header row; for example, the data center feeds are a plain list of CIDR ranges.
File naming conventions
Most feeds put a date in the file name, in YYYYMMDD format. The data center feeds and the weekly user-agent feed use a week number instead.
Rules that apply to the names:
- Dated names carry a date only, with no time or version. A feed that updates more than once a day publishes several files under the same date, and a newer file replaces an older one with the same name.
- Don’t use the file name to find the latest file. Use the last-modified time.
- Week numbers restart at 1 at the start of each year and aren’t zero-padded:
DatacenterSubnetsWeek1throughDatacenterSubnetsWeek9, thenDatacenterSubnetsWeek10onward.
Poll for new files
Check for new files on a schedule that matches the feed’s update frequency. Each check is a last-modified request like the one in Download over FTP. Download a file only when it’s newer than the one you have.
Hourly feeds
Daily feeds
Weekly feeds
- Start checking at the top of the hour.
- Check every 15 minutes.
- Download when the new file is available.
- Stop checking once you’ve found it, and start again at the next hour.
Polling less often than this still works, but you pick up new entries later. At a minimum, check the hourly feeds once every hour.
Load each new file
- Replace the whole list. Replace the old file’s contents with the new file; don’t append to them. If you don’t load the latest list, invalid traffic that Pixalate has since flagged gets through.
- Plan for hourly updates. For hourly feeds, make sure your bandwidth, storage and ingestion jobs can handle a new file every hour.
Check that your data is fresh
- Compare last-modified times. Record the last-modified time of each file you load. If it’s older than the feed’s update frequency allows, your poller has missed a file or the file is late.
- Watch for cadence changes. Changes to update frequency, file names or schemas are announced in the changelog.
- Report missing files. If a file hasn’t arrived when you expect it, contact support@pixalate.com.
Typical file sizes
Plan storage and transfer time for files up to these sizes (as of November 2025):
Related
- Pre-bid data feeds: which feed blocks what, and how often each is updated.
- Blocking FAQ
- Fraud API: real-time lookups instead of lists.